Chronic neck & shoulder pain
When the neck is driving or contributing to shoulder symptoms.
The shoulder is the most mobile joint in the body, which makes it the joint with the most to lose. Protect it for long enough and the range quietly shrinks — until the limit you feel is no longer the limit the joint has.
What is chronic shoulder pain?
Chronic shoulder pain is shoulder pain lasting three months or longer. Common contributors include rotator cuff tendinopathy, subacromial pain, frozen shoulder, and osteoarthritis — but in long-standing cases the pain is frequently maintained by nervous system sensitization and by protective guarding rather than by ongoing tissue damage.
Most chronic shoulder pain starts with something identifiable — an overload, an awkward lift, a period of unusual activity, or a gradual onset with no clear trigger at all. What keeps it going for months is usually a different question from what started it.
The most common category, covering tendinopathy, partial tears, and what used to be called impingement. Pain is typically felt over the outer upper arm, worse with overhead reaching and with lying on that side at night. Terminology has shifted deliberately toward rotator cuff–related shoulder pain, because the older labels implied a mechanical pinching model that the evidence has not supported.
A distinct condition in which the joint capsule thickens and contracts, producing genuine restriction — most characteristically a loss of external rotation that is present whether you move the arm yourself or someone else moves it for you. It is more common in people with diabetes and thyroid disease, and it typically runs through painful, stiff, and thawing phases over one to three years.
Less common in the shoulder than in the knee or hip, and as elsewhere, radiographic severity correlates poorly with symptoms.
The neck refers pain into the shoulder region often enough that it should always be assessed. See chronic neck and shoulder pain.
This is the one most likely to be missed. After months of pain, the nervous system may amplify signals from the region, and the shoulder is held in a protective pattern that itself becomes painful. At this point the pain has less to do with the original tissue problem than with central sensitization and fear of movement.
Because shoulder imaging is remarkably good at finding things that were never causing symptoms. Studies scanning people without any shoulder pain find rotator cuff tears at high rates, rising steeply with age — by the seventh and eighth decades, a substantial fraction of entirely pain-free shoulders have a tear.
This does not make imaging useless. It makes it a source of findings that need interpretation rather than answers that stand alone. A tear may be relevant, incidental, or somewhere in between, and only the clinical picture can settle which.
If you have been shown a scan and told your shoulder is 'torn' or 'degenerating', it's worth asking a specific question: how many people my age, with no pain at all, would have the same finding? The answer is frequently 'a great many'.
The unintended consequence is well documented: alarming imaging language increases fear, fear increases guarding, and guarding reduces the movement that would help. The scan can become part of the problem it was meant to explain.
Outside of true frozen shoulder, most restriction in long-standing shoulder pain is not a mechanical block. It is the point at which the movement is expected to hurt. That distinction is testable, and it matters enormously for treatment.
Karuna Labs ran a small institutional review board–approved mechanistic study in 17 adults with chronic shoulder pain, using virtual reality to test how visual feedback affects pain-free range of motion. Participants moved their arms while a virtual avatar reflected the movement — sometimes matching the arm actually moving, sometimes displaying the movement of the healthy arm as though it were the painful one.
When people moved their unaffected arm but saw the painful side moving, pain-free range of motion significantly decreased for shoulder flexion and scaption. Nothing about the healthy arm had changed. What changed was which shoulder the brain believed was moving.
A healthy shoulder can be made to move less simply by convincing the brain it is the painful one. The limit, in other words, is partly a belief — and beliefs can be revised.
This was single-session mechanistic work in a small sample, not a demonstration of treatment benefit. What it supports is the principle: if perception can contract range of motion, then perception is a legitimate target for expanding it. That principle underpins mirror therapy, graded motor imagery, and our virtual reality training.
The shoulder is a particularly good fit for these approaches, because it is one of a pair. Mirror-based techniques that cannot help the lower back apply directly here.
First-line in essentially every guideline, and the treatment with the most consistent support for rotator cuff–related shoulder pain. The principle is progressive loading within tolerance — enough to drive adaptation, not so much as to provoke repeated flares. Specific exercise programmes have performed comparably to surgery in trials.
Where movement is limited by anticipated pain, graded exposure to the feared movements is the direct treatment. See kinesiophobia.
Corticosteroid injection can provide short-term relief and is sometimes used to create a window in which rehabilitation becomes possible. Benefits are generally short-lived, and injections work best as an adjunct to exercise rather than a replacement for it.
The evidence here has shifted substantially. Placebo-controlled trials of subacromial decompression — including the CSAW and FIMPACT trials — found no meaningful benefit over sham surgery for subacromial shoulder pain. Surgery retains a clear role in specific situations: acute traumatic cuff tears in younger people, instability, and advanced arthritis. It is much less clearly indicated for long-standing non-traumatic shoulder pain.
Management differs. Pain relief and gentle range-of-motion work dominate the painful phase, with more assertive mobilisation as it settles. Hydrodilatation and corticosteroid injection are used, and most cases resolve over one to three years — though residual restriction is common.
If your shoulder pain has lasted months and nobody has asked about sleep, stress, fear of movement, or what you have stopped doing, an important part of the assessment is missing.
Most chronic shoulder pain is not dangerous. Some presentations need prompt assessment, and this list is not something to self-manage around.
New shoulder pain with chest tightness, breathlessness, sweating, or nausea should be treated as a possible cardiac event. Seek emergency care immediately rather than waiting to see whether it settles.
Shoulder pain is coded M25.51- — M25.511 for the right shoulder, M25.512 for the left, and M25.519 for an unspecified shoulder.
Where a specific cause is identified, a more precise code is used: M75.- covers rotator cuff conditions and adhesive capsulitis, and M19.- covers shoulder osteoarthritis. G89.4 (chronic pain syndrome) may be added alongside the site code where chronic pain syndrome is present — see chronic pain syndrome.
By definition it has already lasted three months or more. Without treatment aimed at capacity and sensitivity, shoulder pain can persist for years — a substantial proportion of people still report symptoms 12 to 18 months after first seeking care.
That statistic reflects what usually happens, not what has to. Graded loading and addressing fear of movement change the trajectory for many people.
Move it, within tolerance. Prolonged rest reduces capacity, increases stiffness, and reinforces the nervous system's conclusion that the shoulder is fragile.
This does not mean pushing through severe pain. The aim is a dose that is challenging but settles within about 24 hours. A physiotherapist can help identify that starting point — which is genuinely difficult to judge alone.
No. Many rotator cuff tears — particularly degenerative ones in people over 50 — are managed well without surgery, and exercise programmes have produced comparable outcomes to surgical repair in trials of degenerative tears.
Surgery is more clearly indicated for acute traumatic tears in younger, active people, and where there is significant weakness or a failed course of good rehabilitation. This is a decision for you and your surgeon, informed by your symptoms rather than by the scan alone.
Night pain is characteristic of rotator cuff–related shoulder pain and frozen shoulder. Lying down loads the shoulder differently, there is less distraction, and inflammatory pain often follows a daily rhythm.
Practical measures — sleeping on the other side, supporting the arm on a pillow, avoiding lying directly on the shoulder — help. Persistent severe night pain, particularly with other red flags, should be assessed.
Many people achieve substantial reduction in pain and return to full activity. Whether 'cured' is the right word depends on the driver — frozen shoulder typically resolves over time, while rotator cuff–related pain is usually better framed as building capacity than as repairing something.
Where sensitization and guarding are maintaining the pain, treatment aimed at the nervous system can produce large improvements even after years of symptoms.
Talk with our care team about your pain, your history, and whether KVET™ is right for you — free, and from the comfort of home.